The proposed study will examine patterns of contraceptive method switching among both married and unmarried women in the United States. Based on data from Cycle III of the National Survey of Family Growth, one-year switching rates for individual contraceptive methods will be estimated with a hazards model life table approach. Using these switching rates, transition matrices will be constructed to: (1) examine the destination distribution of women with a given method origin; and (2) examine the origin distribution of women with a given method destination. The effects of individual characteristics on method switching will also be investigated, permitting the identification and comparison of subgroup differences in contraceptive switching processes. The proposed study will significantly expand current knowledge of contraceptive switching and its effects on method utilization.